基于信息熵与遗传算法的公共交通线路选择模型  

Transit Route Choice Model Based on Information Entropy and Genetic Algorithm

作　　者：刘倩茜[1] 高宁波[1] 郑丽媛[1] 吴鹏[1] 

机构地区：[1]南京理工大学,自动化学院,南京210094

出　　处：《交通运输工程与信息学报》2016年第1期87-94,共8页Journal of Transportation Engineering and Information

基　　金：国家自然科学基金资助项目(51178157);教育部人文社会科学研究项目(NO.12YJCH071);国家统计科研计划项目(NO.2012LY150);江苏省高校“青蓝工程”资助项目(NO.201211)

摘　　要：针对公共交通线路选择问题,旨在分析起讫点间公共交通运行的时间和线路类型等级对居民公共交通线路选择行为的影响,在此基础上根据路权等级划分城市公共交通方式的类型并运用遗传算法仿真求出最短路径。利用信息熵理论和多属性综合决策方法对无偏好下居民选择公共交通线路的不确定性问题进行解析,搭建路径综合属性值的计算模型,获得每条路径的综合属性值,并基于最大值原理获得最优出行路径。最后,将模型运用于郑州市金水区某个公共交通网络出行实例中。结果表明:在无偏好下,线路5的综合属性值为0.999,而按照直观的最短路线路1的综合属性值仅为0.884,可见,线路类型等级对公共交通线路选择具有显著的影响。Aimming at route choice of public transit, the impact of public transit operation time and travel mode between the origin-destination pair on the resident＇s choice behavior was analyzed. Dividing the urban transit types according to the right of way level and using the genetic algorithm, the shortest path was simulated. Taking the advantage of the information entropy theory and the multi-property integrated decision method for residents to choice the public transit routes without preference, the problem＇s uncertainty was resolved. A calculative model of path comprehensive property value was established, and the comprehensive property values were obtained for each path; then, the best travel route was get based on the maximum principle. Finally, the model was applied to a public traffic travel network of Jinshui district of Zhengzhou City. The results showed that： in the absence of preference, the comprehensive property value of line 5 was 0.999, but in accordance with visualization, the comprehensive property value was only 0.884. So, we get that line type level has a significant effect on the public transit line selection.

关 键 词：交通工程 城市交通 公共交通 信息熵 遗传算法 

分 类 号：U491.17[交通运输工程—交通运输规划与管理]